• Aucun résultat trouvé

Potential drug-drug interactions and adverse drug reactions in patients with liver cirrhosis

N/A
N/A
Protected

Academic year: 2021

Partager "Potential drug-drug interactions and adverse drug reactions in patients with liver cirrhosis"

Copied!
10
0
0

Texte intégral

(1)

PHARMACOEPIDEMIOLOGY AND PRESCRIPTION

Potential drug-drug interactions and adverse drug reactions

in patients with liver cirrhosis

Carmen C. Franz&Sabin Egger&Christa Born& Alexandra E. Rätz Bravo&Stephan Krähenbühl

Received: 4 May 2011 / Accepted: 21 July 2011 / Published online: 13 August 2011 # Springer-Verlag 2011

Abstract

Background and aims Patients with liver cirrhosis may be at risk for potential drug-drug interactions (pDDIs) and/or adverse drug reactions (ADRs) due to the severity of their disease and comorbidities associated with polypharmacy. Methods We performed a cross-sectional retrospective study including 400 cirrhotic patients and assessed diagnoses, medication patterns, pDDIs, and ADRs at hospital admission.

Results The median (range) age of the patients was 60 (21– 88) years; 68.5% were male. They had a total of 2,415 diagnoses, resulting in 6 (1–10) diagnoses per patient. Frequent were diagnoses of the digestive system (28.4%), circulatory system (14.2%), blood and blood-forming organs (8.7%), and psychiatric disorders (7.5%); 60.7% of the diagnoses were not liver-associated. The median number of drugs per patient was 5 (0–18), whereof 3 (0– 16) were predominantly hepatically eliminated. Drugs were primarily indicated for gastrointestinal, cardiovascular, or nervous system disorders, reflecting the prevalent diagno-ses. In 112 (28%) patients, 200 ADRs were detected, mainly associated with spironolactone, torasemide, furose-mide, and ibuprofen. In 86 (21.5%) patients, 132 pDDIs

were detected. Seven of these pDDIs were the direct cause of 15 ADRs, whereof 3 resulted in hospital admission. Patients with ADRs were older, had more comorbidities, were treated with more drugs, and had a worse renal function and more pDDIs than patients without ADRs. Conclusions Pharmacotherapy is complex in cirrhotic patients. Hepatologists should know the principles of dose adjustment in cirrhosis and renal failure, but also the most important pDDIs of the drugs used to treat liver disease and comorbidities in this population.

Keywords Liver cirrhosis . Drug-drug interactions . Adverse drug reactions . Dose adjustment

Abbreviations

pDDIs Potential drug-drug interactions

ADRs Adverse drug reactions

NSAIDs Nonsteroidal anti-inflammatory drugs

Q0 Extrarenal elimination fraction

ATC code Anatomical Therapeutic Chemical Classifica-tion System

ACE Angiotensin-converting enzyme

HSCT Hematopoietic stem cell transplantation

RAAS Renin angiotensin aldosterone system

SSRI Selective serotonine reuptake inhibitor

COX Cyclooxygenase

Introduction

Liver cirrhosis remains a frequent cause of morbidity and mortality in most countries, including countries in Europe. Between 1997 and 2001, the yearly mortality rate due to liver cirrhosis was between 9.7 (Netherlands) and 43.5

Electronic supplementary material The online version of this article (doi:10.1007/s00228-011-1105-5) contains supplementary material, which is available to authorized users.

C. C. Franz

:

S. Egger

:

C. Born

:

A. E. Rätz Bravo

:

S. Krähenbühl (*)

Division of Clinical Pharmacology and Toxicology, University Hospital,

4031 Basel, Switzerland e-mail: kraehenbuehl@uhbs.ch A. E. Rätz Bravo

Regional Pharmacovigilance Center, University Hospital, Basel, Switzerland

(2)

(Austria) per 100,000 males and between 5.6 (Sweden) and 16.7 (Austria) per 100,000 females [1].

Since the liver plays a crucial role in the metabolism of endogenous and exogenous substances, impaired hepatic function may influence the pharmacokinetics of drugs used in cirrhotic patients. The absorption process may be altered [2,3] and bioavailability may be increased due to porto-systemic shunting [2, 4]. The free fraction and possibly also the free concentration of highly protein-bound drugs is increased in patients with hypoalbumine-mia [2]. Finally, hepatic drug clearance is usually decreased due to lower hepatic blood flow [2, 5] and decreased activity of phase I enzymes [2,6, 7]. Pharma-codynamic changes are also prevalent in patients with liver cirrhosis. Increased sensitivity has been shown for central effects of morphine [8] and benzodiazepines [9] and for renal adverse effects of nonsteroidal anti-inflammatory drugs (NSAIDs) [10].

All of these factors can potentially influence the effectiveness of a drug and/or the likelihood that a drug is causing adverse reactions. Adverse drug reactions (ADRs) may further increase morbidity and mortality in patients with liver disease.

The current study had several aims concerning drug treatment of patients with liver cirrhosis. First, we wanted to find out which drugs are commonly prescribed in this group of patients. Secondly, we investigated the quantity and severity of potential drug-drug interactions (pDDIs) in these patients. Thirdly, we identified the ADRs. For this purpose, we characterized the medication pattern of 400 patients with liver cirrhosis and assessed the prevalence of pDDIs and ADRs at hospital admission.

Methods Patients

In the present cross-sectional, retrospective study, we included 400 patients with liver cirrhosis diagnosed by liver histology and/or typical clinical, sonographic, and computer tomographic signs. They were hospitalized at the University Hospital, Basel, Switzerland, between January 2002 and December 2007. The protocol of the study was accepted by the cantonal Ethics Committee.

Data collection

For each patient, demographic and clinical data, diagnoses, drugs administered, characteristics of the drugs adminis-tered (dosage and extrarenal elimination fraction, Q0), and

pDDIs and ADRs [11] were collected at hospital admission. Creatinine clearance was calculated by the Cockcroft Gault

equation [12]. Severity of liver cirrhosis was classified by the Child Pugh Score [13]. Drugs were grouped according to the Anatomical Therapeutic Chemical Classification System (ATC code). Drugs with a Q0≥0.5 were defined as

primarily hepatically eliminated. Potential DDIs were determined by screening the drug profiles using the online version of DRUG-REAX (Micromedex® 1.0 Healthcare Series, http://www.micromedex.com). Only pDDIs with moderate or major severity were considered. All ADRs were classified with a definite, probable, or possible causality rating as described previously [14].

Statistical analysis

The data were descriptively analyzed using Excel and/or SPSS (version 15.0). Comparisons between patients with ADRs and those without ADRs were performed using Student’s t-test or the chi-squared test without correction for repetitive testing. A significance level of 5% was chosen.

Results

Patient characteristics

All patients studied were adults with males being more prevalent than females (supplementary Table 1). Most patients were in the Child Pugh classes B and C. The most frequent cause of liver cirrhosis was alcohol (69.8%), followed by viral hepatitis (13.5%) or a combination of both (9.7%). Almost 20% of the patients died during hospitalization, reflecting the severity of this disease.

The patients had a total number of 2,415 diagnoses at hospital admission, resulting in a median number of 6 (1– 10) diagnoses per patient (supplementary Table 2). Most common were diseases of the digestive (28.4% of all diagnoses) or circulatory system (14.2%) as well as diseases of the blood and blood-forming organs (8.7%) and psychiatric disorders (7.5%). Approximately 40% of all diagnoses were associated with liver cirrhosis, e.g., spon-taneous bacterial peritonitis, esophageal varices, and variceal bleeding.

Medication at hospital admission

The patients had a total of 1,999 drugs at hospital admission (Table 1). The median number of drugs per

patient was 5 (0–18); a median of 3 (0–16) were

predominantly hepatically eliminated. Most prevalent were drugs affecting the alimentary tract and metabolism, mainly vitamins and proton pump inhibitors, as well as drugs for

(3)

Table 1 Drugs at hospital admission for patients with liver cirrhosis (n=400 patients)

Number of drugs (% of patients receiving corresponding drug)

Number of drugs with Q0≥ 0.5a(% of patients receiving corresponding drug)

Number of drugs at hospital admission 1,999 1,360

Drugs per patient at hospital admissionb 5 (0–18) 3 (0–16)

Alimentary tract and metabolismc 650 255

Vitaminsc 207 6

Thiamine 89 (22.3%) 0

Proton pump inhibitors 154 (38.5%) 154 (38.5%)

Osmotically acting laxatives 83(20.8%) 0

Blood glucose–lowering drugs (excl. insulins)c 65 38

Magnesium 33 (8.3%) 0

Insulins and analogues 27 (6.8%) 27 (6.8%)

Calcium 27 (6.8%) 0

Propulsives 12 (3.0%) 12 (3.0%)

Potassium 11 (2.8%) 0

Cardiovascular systemc 633 532

Potassium-sparing diureticsd 160 (40.0%) 160 (40.0%)

Loop diuretics (high ceiling) 157 (39.3%) 127 (31.8%)

Betablockers 146 (36.5%) 134 (33.5%)

Propranolol 78 (19.5%) 78 (19.5%)

ACE inhibitors 34 (8.5%) 8 (2.0%)

Calcium antagonists 25 (6.3%) 25 (6.3%)

Statins 23 (5.8%) 23 (5.8%)

Angiotensin receptor blockers 19 (4.8%) 19 (4.8%)

Thiazides 19 (4.8%) 0

Organic nitrates 12 (3.0%) 12 (3.0%)

Amiodarone 7 (1.8%) 7 (1.8%)

Nervous systemc 270 257

Benzodiazepines and related drugs 102 (25.5%) 102 (25.5%)

Lorazepam 28 (7.0%) 28 (7.0%) Zolpidem 27 (6.8%) 27 (6.8%) Oxazepam 14 (3.5%) 14 (3.5%) Diazepam 13 (3.3%) 13 (3.3%) Opioids 58 (14.5%) 58 (14.5%) Methadone 29 (7.3%) 29 (7.3%)

Antidepressants, excl. SSRIe 30 (7.5%) 30 (7.5%)

SSRI 21 (5.3%) 21 (5.3%)

Neuroleptics 17 (4.3%) 15 (3.8%)

Antiepileptics 16 (4.0%) 10 (2.5%)

Dopaminergic agents 8 (2.0%) 8 (2.0%)

Blood and blood-forming organsc 154 96

Phytomenadione 44 (11.0%) 44 (11.0%)

Platelet aggregation inhibitors (incl. aspirin low dose) 30 (7.5%) 30 (7.5%)

Iron 26 (6.5%) 0 Oral anticoagulants 21 (5.3%) 21 (5.3%) Heparins 15 (3.8%) 2 (0.5%) Folic acid 13 (3.3%) 0 Musculoskeletal systemc 93 89 NSAIDs 28 (7.0%) 28 (7.0%) Paracetamol 23 (5.8%) 23 (5.8%)

(4)

the cardiovascular system, primarily potassium-sparing diuretics, loop diuretics, and betablockers. Approximately 10% of all patients were treated with an angiotensin-converting enzyme (ACE) inhibitor. The most frequent drugs for the nervous system were benzodiazepines and benzodiazepine-related drugs as well as opioids. Eleven percent of the patients were treated with phytomenadione, 7.5% with platelet-aggregation inhibitors, and 5% with oral anticoagulants. Astonishingly, 11% of the patients were treated with a COX inhibitor (NSAIDs, analgesic aspirin, or COX-2 inhibitor).

About 68% of all administered drugs were eliminated primarily hepatically (Q0≥0.5).

pDDIs and ADRs at hospital admission

In 21.5% of all patients, a median of 1 (1–5) pDDI was detected (Table 2). Most prevalent possible adverse reac-tions due to pDDI were hyperkalemia (potassium-sparing diuretics, ACE inhibitors, and potassium chloride), hypo-glycemia (betablockers combined with insulin, sulfonylureas, and/or repaglinide), increased bleeding risk (anticoagulants such as dalteparin or phenprocoumon combined with NSAIDs), respiratory depression (benzodiazepines combined with opiates or phenobarbital), and cardiac problems (cardiac

depression, QT prolongation). Of all pDDIs, 12.9% resulted in an ADR.

ADRs were detected in 28% of the patients at entry (Table 3). Relative to the number of patients in each Child Pugh class, patients in class Child Pugh A were more frequently affected by ADRs (35.7% of patients) than those in class Child Pugh B (26.1%) or Child Pugh C (26.6%). Nonetheless, most ADRs (43.0%) occurred in patients with liver cirrhosis Child Pugh C. The drugs most frequently associated with an ADR were spironolactone, torase-mide, furosetorase-mide, and ibuprofen. Most frequently, ADRs resulted in metabolic disorders (mainly hyper-kalemia or hyponatremia associated with diuretics and/or ACE inhibitors), in gastrointestinal bleeding (associated with the use of NSAIDs or oral anticoagulants), and in urinary system disorders (mainly worsening renal func-tion due to the use of diuretics and/or ACE inhibitors). Five percent of all ADRs affected the liver and/or the biliary system.

Sixteen ADRs (8%) were the cause of hospital admis-sion. These ADRs consisted of gastrointestinal bleeding associated with low dose aspirin, ibuprofen, or phenpro-coumon; hyperkalemia associated with spironolactone or perindopril; and worsening renal function or ascites accumulation associated with ibuprofen. Fifteen ADRs

Table 1 (continued)

Number of drugs (% of patients receiving corresponding drug)

Number of drugs with Q0≥ 0.5a(% of patients receiving corresponding drug)

Allopurinolf 14 (3.5%) 14 (3.5%)

Aspirin, analgesic 8 (2.0%) 8 (2.0%)

COX-2 inhibitors 7 (1.8%) 7 (1.8%)

Respiratory systemc 75 43

Anti-infectives for systemic usec 63 30

Antiviralsc 38 20

Antibacterials 25 (6.3%) 10 (2.5%)

Fluoroquinolones 9 (2.3%) 4 (1.0%)

Systemic hormonal preparations, excl. sex hormones and insulinsc 31 31

Corticosteroids 16 (4.0%) 16 (4.0%)

Thyroid hormones 13 (3.3%) 13 (3.3%)

Antineoplastic and immunomodulating agentsc 15 10

Genito-urinary system and sex hormonesc 10 9

Variousc 5 5

a

Only drugs with known Q0included b

Data are presented as median (range) c

One individual patient may have >1 drug of the corresponding group, % not calculated d

Spironolactone accounts for 97.5% of this group. It is mainly converted to active metabolites (the two major ones being canrenone and 7-alpha-thiomethylspironolactone), which are primarily renally eliminated

eIncluding tri-, tetracyclic antidepressants, and venlafaxine

fAllopurinol is rapidly converted by the liver to the slightly less active oxypurinol, which is renally eliminated. Dosage adjustment is necessary in both patients with liver and renal insufficiency

(5)

(7.5%) resulted from a DDI; among them five patients with bleeding disorders (gastrointestinal bleeding, epistaxis, anemia), four patients with hyperkalemia, three patients with cardiovascular disorders (hypotension, bradycardia, torsade de pointes) as well as one patient each with a psychiatric disorder, somnolence, and collapse. Three DDI-associated ADRs were the reason for hospital admission, namely gastrointestinal bleeding due to the combination of aspirin (100 mg/day) and phenprocoumon, symptomatic bradycardia due to the combination of amiodarone and propranolol, and hyperkalemia due to the combination of spironolactone and perindopril.

The 36 NSAIDs prescribed (NSAIDs and analgesic aspirin) were associated with 24 ADRs in 18 patients (50% of all patients with a NSAID) (Fig. 1). The most prevalent ADRs due to NSAIDs were gastrointestinal bleeding (14/24), bleeding-associated anemia (3/24), exacerbation of ascites (2/24), and thrombocytopenia (2/24). The 53 ACE inhibitor or sartan prescriptions resulted in 20 ADRs in 10 patients (19% of patients treated with an ACE inhibitor or sartan, Fig. 1). Symptoms observed were worsening renal function (6/ 20), syncope (4/20), hyperkalemia (3/20), hyponatremia (3/20), and hypotension (2/20). In contrast, of 146

Table 2 Major and moderate pDDIs in patients with liver cirrhosis (n=400 patients)

DDI/potential outcome Number of pDDIs Interacting drugs (number of cases) Total pDDIs 132 (100%)

Major pDDIs 56 (42.1%) Moderate pDDIs 76 (57.9%) Number of different pDDIs 91 Number of patients

with≥1 pDDI

86 (21.5% of all patients) pDDIs per patienta, b 1 (1–5)

DDIs resulting in an ADR 17 (12.9% of all pDDIs)

Hyperkalemia 24 (18.2%) Major: Potassium-sparing diuretics + ACE inhibitors (9); potassium chloride + spironolactone (4), potassium chloride + lisinopril (1) + risk for nephrotoxicity 9 (6.8%) Moderate: Spironolactone + valsartan (1)

Moderate: Potassium-sparing diuretics + NSAID (9)

Hypoglycemia 23 (17.4%) Moderate: Betablocker + insulin (14), betablocker + sulfonylureas (8), betablocker + repaglinide (1)

Increased bleeding risk 17 (12.9%) Major: Dalteparin + acetylsalicylic acid (low dose) (1), dalteparin +

clopidogrel (1), dalteparin + phenprocoumon (1), dalteparin + ibuprofen (1), acetylsalicylic acid (low dose) + phenprocoumon (2),

acetylsalicylic acid (low dose) + venlafaxine (1)

Moderate: Phenprocoumon + allopurinol (3), phenprocoumon + amiodarone (3), phenprocoumon + diclofenac (1), acetylsalicylic acid (low dose) + verapamil (2), acetylsalicylic acid (low dose) + ibuprofen (1)

Respiratory depression 10 (7.6%) Major: Benzodiazepines + opiates (7), benzodiazepines + phenobarbital (2), fentanyl + hydrocodone (1)

Cardiac depression 9 (6.8%) Major: Betablocker + calcium antagonist (4), betablocker + amiodarone (2) Moderate: Digoxin + betablocker (3)

QT prolongation 7 (5.3%) Major: Amitriptyline + sulfamethoxazole, trimethoprim (2), ciprofloxacin + propafenone (1), fluoxetine + haloperidol (1), fluoxetine + methadone (1), risperidone + tramadol (1), trimipramine + venlafaxine (1)

Digoxin toxicity 6 (4.5%) Major: Digoxin + hydrochlorothiazide (2),

digoxin + spironolactone (2), digoxin + amiodarone (1) Moderate: Digoxin + furosemide (1)

Altered methadone exposure 6 (4.5%) Moderate: Methadone + HIV protease inhibitor (4), methadone + efavirenz (2) Serotonin syndrome 4 (3.0%) Major: Mirtazapine + fluoxetine (1), mirtazapine + tramadol (1),

mirtazapine + venlafaxine (1), tramadol + venlafaxine (1) Reduced efficacy of levodopa 4 (3.0%) Moderate: Levodopa + iron (2), levodopa + levomepromazine (1),

levodopa + olanzapine (1)

Other 22 (16.7)

aReferring to the patients with one or more pDDI (n=86) bData are presented as median (range)

(6)

Table 3 Prevalence of ADRs with a definite, probable, or possible causality rating in 400 patients with liver cirrhosis

ADR Number of ADRs (total; according to

Child Pugh A, B, C)

Drugs associated with ADRsa(cases according to Child Pugh A, B, C)b

Total ADR 200 (100.0%) Spironolactone (9, 11, 30), torasemide (5, 7, 20), furosemide (4, 1, 10),

ibuprofen (1, 4, 6)

Child Pugh A 46 (23.0%)

Child Pugh B 68 (34.0%)

Child Pugh C 86 (43.0%)

Number of ADR per patientc,d 1 (1–5)

Child Pugh A 1 (1–5)

Child Pugh B 1 (1–5)

Child Pugh C 1.5 (1–5)

Patients with≥1 ADR 112 (28% of all patients)

Child Pugh A 25 (35.7% of Child Pugh A patients)

Child Pugh B 41 (26.1% of Child Pugh B patients)

Child Pugh C 46 (26.6% of Child Pugh C patients)

ADR with definite/probable causality rating

24 (12.0%) (8, 11, 5) Spironolactone (3, 1, 1), phenprocoumon (0, 4, 0),

ibuprofen (0, 1, 3), acetylsalicylic acid low dose (1, 3, 0) ADRs with possible causality

rating

176 (88.0%) (38, 57, 81) Spironolactone (6, 10, 29), torasemide (3, 7, 19),

furosemide (3, 1, 10), propranolol (2, 2, 3) ADRs as a reason for

hospital admission

16 (8.0%) (6, 5, 5) Spironolactone (3, 0, 1), acetylsalicylic acid low dose (1, 2, 0),

ibuprofen (0, 0, 3), torasemide (2, 0, 1), perindopril (2, 0, 0), phenprocoumon (0, 2, 0)

ADRs due to≥1 DDI 15 (7.5%) (6, 7, 2) Major: Spironolactone + ACE inhibitors (1, 0, 1), opiates + benzodiazepines

(1, 1, 0), acetylsalicylic acid low dose + dalteparin (0, 1, 0), acetylsalicylic acid low dose + phenprocoumon (0, 1, 0), dalteparin + ibuprofen (1, 0, 0), diltiazem + betablockers (2, 0, 0), amiodarone + propranolol (0, 1, 0) ADR due to DDI causing

hospital admission

3 (2.0%) (1, 2, 0) Acetylsalicylic acid low dose + phenprocoumon (0, 1, 0),

perindopril + spironolactone (1, 0, 0), amiodarone + propranolol (0, 1, 0) ADR according to system organ class

Metabolic and nutritional disorders

54 (27.0%) (13, 13, 28) Spironolactone (4, 6, 17), torasemide (1, 2, 6), furosemide (3, 0, 3),

hydrochlorothiazide (1, 2, 0), chlortalidone (3, 0, 0), amiloride (0, 3, 0), ramipril (0, 0, 2), enalapril (0, 2, 0) Gastro-intestinal system

disorders

30 (15.0%) (6, 13, 11) Acetylsalicylic acid low dose (1, 4, 0) and analgesic (1, 1, 1),

ibuprofen (0, 2, 3), mefenamic acid (1, 2, 1), iron (0, 2, 1), spironolactone (0, 0, 3), phenprocoumon (0, 2, 0)

Urinary system disorders 25 (12.5%) (3, 9, 13) Torasemide (2, 3, 8), spironolactone (2, 2, 6), furosemide (0, 1, 4),

enalapril (0, 1, 1), lisinopril (1, 0, 1) Cardiovascular disorders,

general

16 (8.0%) (6, 3, 7) Furosemide (1, 0, 2), amlodipine (2, 0, 1), diltiazem (2, 0, 0),

ramipril (0, 1, 1), spironolactone (1, 0, 1), torasemide (0, 0, 2) Central and peripheral

nervous system disorders

15 (7.5%) (1, 6, 8) Zolpidem (1, 0, 2), oxazepam (0, 1, 1), propranolol (0, 0, 2),

ropinirole (0, 2, 0), spironolactone (0, 1, 1), torasemide (0, 1, 1) Liver and biliary system

disorders

10 (5.0%) (1, 3, 6) Spironolactone (1, 1, 1), enalapril (0, 1, 1)

Psychiatric disorders 10 (5.0%) (4, 4, 2) Midazolam (1, 0, 1), oxazepam (0, 2, 0), propranolol (1, 1, 0),

zolpidem (1, 0, 1) Platelet, bleeding and

clotting disorders

10 (5.0%) (3, 5, 2) Spironolactone (1, 0, 1), torasemide (1, 0, 1)

Red blood cell disorders 8 (4.0%) (2, 2, 4) Ibuprofen (0, 1, 2), torasemide (1, 0, 2)

Heart rate and rhythm disorders

4 (2.0%) (1, 2, 1) -e

Other 18 (9.0%) (6, 8, 4) -e

a

Most frequent drugs associated with ADR mentioned b

Sum of cases may exceed the number of ADRs (more than one drug can cause the same ADR) cReferring to the patients with one or more ADRs (n=112)

dData are presented as median (range) e

(7)

betablocker prescriptions, only 12 ADRs (hypotension, syncope, confusion) were identified in nine patients (6% of all patients with a betablocker, Fig.1).

Cirrhotic patients with one or more ADR compared with cirrhotic patients without an ADR

When comparing patients with ADRs to those without ADRs (Table 4), patients with ADRs were significantly

older than those without ADRs (61 vs. 58 years, p < 0.05), had a lower creatinine clearance (64.3 vs. 91.6 ml/min, p < 0.001), had more total diagnoses (6.38 vs. 5.91, p < 0.05), as well as more non-liver-associated diagnoses (4.20 vs. 3.77, p<0.05). They had more drugs prescribed (6 vs. 4, p< 0.001), as well as more drugs with predominantly hepatic elimination (4 vs. 3, p < 0.001). Patients with ≥1 hepatically eliminated drug were more prevalent in the ADR group (99.1 vs. 81.6%, p < 0.001). The same was true NSAIDs

Prescriptions Total number of ADRs

Patients with ADR

Number (n) 0 5 10 15 20 25 30 35 40 ACEIs or ARBs Prescriptions Total number of ADRs

Patients with ADR

Number (n) 0 5 10 15 20 25 30 35 40 45 50 55 60 Betablockers Prescriptions Total numberof ADRs

Patients with ADR

Number (n) 0 20 40 60 80 100 120 140 160

Fig. 1 Number of prescriptions, adverse drug reactions (ADRs), and patients with ADRs for specific drug classes. NSAIDs Nonsteroidal anti-inflammatory drugs, ACEIs angiotensin-converting enzyme inhibitors, ARBs angiotensin receptor blockers

Table 4 Cirrhotic patients with one or more ADRs in comparison with cirrhotic patients without an ADR

Characteristics Patients with ADR (n=112) Patients without ADR (n=288) p-value

Age (years)a 61 (35–88) 58 (21–87) 0.017

Male 77 (68.8%) 197 (68.4%) 0.947

Creatinine clearance (ml/min)a, b 64 (9–290) 92 (9–280) 0.001

BMI (kg/m2)a, c 24.3 (16.0–42.0) 24.9 (13.5–47.2) 0.997

Child Pugh classification 0.282

Child Pugh A 25 (22.3%) 45 (15.6%)

Child Pugh B 40 (35.7%) 117 (40.6%)

Child Pugh C 47 (42.0%) 126 (43.8%)

Diagnoses per patienta 6 (3–10) 6 (1–10) 0.036

Drugs per patient 6 (0–15) 4 (0–18) <0.001

Drugs with Q0≥0.5 per patient 4 (0–12) 3 (0–16) <0.001

Patients with≥1 hepatically eliminated drug 111 (99.1%) 235 (81.6%) <0.001

Number of pDDIs per patient 0 (0–5) 0 (0–5) 0.012

Patients with≥1 pDDI 35 (31.3%) 51 (17.8%) 0.004

Length of hospital stay (days)a 12 (1–77) 14 (2–116) 0.984

Patients died during hospitalization 23 (20.5%) 44 (15.3%) 0.233

Cause of cirrhosis 0.056

Alcohol 87 (77.7%) 192 (66.7%)

Viral hepatitis 9 (8.0%) 45 (15.6%)

Both 7 (6.3%) 32 (11.1%)

a

Data are presented as median (range) b

Due to incomplete data (body weight, serum creatinine), n=107 and 279 for patients with and without ADR, respectively c

(8)

for patients with≥1 pDDI at hospital admission (31.3 vs. 17.8%). Furthermore, pDDIs were more prevalent in patients with ADRs than in the control group (0.50 vs. 0.26 per patient, p < 0.05).

Discussion

Studies on patients with liver cirrhosis focusing on drug therapy and drug related problems are scarce in the literature. Lucena et al. investigated prescribing patterns and drug use in patients with liver cirrhosis [15,16]. To prevent or treat complications of cirrhosis, diuretics, anti-ulcer drugs, laxatives, and vitamin K were the drugs prescribed most often [15]. Frequent medications for nonhepatic comorbidities consisted of insulin, oral antidiabetics, cardiovascular drugs (calcium antagonists, ACE inhibitors, angiotensin receptor blockers), as well as drugs for the nervous (anxiolytics, hypnotics) and respiratory system [16]. The medication pattern of the patients in our study was similar to the patients reported by Lucena et al. [15,

16], suggesting that these studies reliably reflect the medica-tion pattern in cirrhotic patients.

Every fifth patient in our study population had a pDDI, every fourth had an ADR, and 8% of the patients were hospitalized due to an ADR. This is in line with the 5–10% prevalence for ADR-related hospitalizations found in the meta-analysis by Lazarou et al. [17], but slightly more than the 5.1–6.5% reported in a retrospective cohort study [18] and in a prospective observational study [19]. Compared to patients without ADRs, patients with ADRs had more diagnoses and more drugs prescribed, received more drugs eliminated hepatically, had more pDDIs, and had a more compromised renal function.

Polypharmacy is a known risk factor for ADRs [11,20,

21] and DDIs [22, 23]. Our data indicate that cirrhotic patients who have more comorbidities have more drugs prescribed and are therefore at a higher risk for ADRs. The relationship between number of diagnoses and number of drugs prescribed is well established [23]. The resulting polypharmacy is a risk factor for pDDIs [22,23] and also for ADRs [11,20,21], which may be related to DDIs.

Our data suggest also that treatment with drugs with predominantly hepatic elimination is a risk factor for ADRs. More than 50% of the drugs used in our study fall into this category. Patient exposure to such drugs may be increased mostly due to elevated oral bioavailability and/or decreased hepatic clearance, possibly leading to an increased incidence of dose-dependent ADRs [2]. Astonishingly, systematic publications focusing on hepatically eliminated drugs as a risk factor for ADRs in patients with liver disease are lacking. The drugs with predominantly hepatic elimination associated with ADRs in our population were mostly cardiovascular drugs (torasemide, spironolactone,

proprano-lol, amlodipine, diltiazem), NSAIDs (ibuprofen, acetylsali-cylic acid, mefenamic acid), phenprocoumon, and benzodiazepines or related agents (midazolam, oxazepam, zolpidem). If possible, such drugs should be started at a low dose with careful up-titration until reaching a satisfactory drug response or toxicity.

A further risk factor for ADRs is impaired renal function. Impaired glomerular filtration is a well known risk factor for ADRs also in other populations such as the elderly [11, 24]. In our study, patients with ADRs had a lower creatinine clearance as compared to patients without ADRs. Impaired glomerular filtration may be associated with decreased renal clearance and increased exposure to drugs with a predominantly renal excretion. Importantly, patients with liver cirrhosis and ascites can have a creatinine clearance <60 ml/min in spite of a normal serum creatinine [25], mostly due to impaired hepatic formation of creatine and increased tubular secretion of creatinine [2,

26]. Since the Cockcroft formula may overestimate the creatinine clearance in cirrhotic patients, drugs with predominantly renal elimination and dose-dependent ADRs should be dosed very carefully in cirrhotic patients [2].

Another important risk factor for ADRs is the presence of pDDIs. In our study, 7.5% of all ADRs were due to a DDI and 12.9% of the DDIs resulted in an ADR. In a recent review of hospitalized patients on different wards, 17% (range 5–31%) of all ADRs were reported to be due to DDIs [11]. In patients with hematopoietic stem cell transplantation (HSCT), 16% of the ADRs were caused by a DDI and 33% of all DDIs resulted in an ADR [14]. A comparison of the findings in patients with liver cirrhosis suggests that the DDIs in cirrhotics are less severe compared to DDIs in patients with HSCT. This is due to the fact that imidazole and triazole antimycotics used in patients with HSCT are CYP inhibitors interacting with immunosuppressants such as cyclosporin and tacrolimus, which are used routinely in HSCT patients. Nevertheless, pDDIs possibly resulting in severe ADRs are present also in cirrhotic patients; they are known and should be avoided.

The drugs most frequently involved in ADRs and pDDIs in our patients were ACE inhibitors, diuretics, NSAIDs, and oral anticoagulants. ACE inhibitors predispose cirrhotic patients to electrolyte disturbances and renal ADRs. The risk for hyperkalemia in cirrhotic patients treated with ACE inhibitors is increased 5.2-fold compared to patients without liver disease [27]. Patients with liver cirrhosis and portal hypertension have an activation of the renin angiotensin aldosterone system (RAAS) and of the sympa-thetic nervous system, leading to renal vasoconstriction, impaired renal perfusion and glomerular filtration [28], and increased sodium retention [29]. Since drugs interfering with the RAAS such as ACE inhibitors or sartans can further impair glomerular filtration due to reduced filtration

(9)

pressure, they should be used very cautiously in cirrhotic patients. Nevertheless, ACE inhibitors and sartans are prescribed frequently in cirrhotic patients [16].

NSAIDs and COX-2 inhibitors block renal production of prostaglandins, possibly leading to impaired renal perfusion and glomerular filtration, sodium retention, and increase in ascites [2, 10]. Furthermore, NSAIDs may be associated with bleeding from esophageal varices and/or gastrointes-tinal ulcers due to their toxic effects on gastrointesgastrointes-tinal epithelia and inhibition of thrombocyte function. In a case-control study including patients with esophageal varices with or without variceal bleeding, the use of NSAIDs in the week prior to the index day was significantly more common in bleeding patients (OR =2.8) [30]. Taking into account the risks for gastrointestinal bleeding, deterioration of renal function and increase in ascites, it is astonishing that 7% of our patients used NSAIDs and 2% analgesic aspirin. A clearer communication of the risks associated with the use of these drugs and of the analgesic alternatives in this population is therefore necessary.

Our study has several limitations. A first limitation is the retrospective character of the study. The elaborated data were therefore limited to the information provided in the medical records, and it was sometimes not possible to obtain more information on the patient’s situation or drug history prior to hospitalization. A second limitation is the limited sample size, which resulted in relatively small numbers of ADRs and DDIs. Nevertheless, we are convinced that the study provides important safety data in patients with liver cirrhosis and helps in identifying medication risks.

From the data of our study, we conclude that patients with liver cirrhosis have many comorbidities predisposing them to polypharmacy, which is associated with pDDIs and ADRs. Besides polypharmacy, important risk factors for ADRs in cirrhotic patients are lack of dose adjustment of drugs eliminated predominantly by the liver or by the kidney and certain pDDIs. Hepatologists should therefore not only know the principles of dose adjustment in patients with liver and/or renal failure, but also the most important DDIs of the drugs used to treat liver disease and comorbidities in this population.

Conflict of interest None of the authors indicates a conflict of interest with this work.

Financial support S.K. is supported by the Swiss National Science Foundation (31003A_132992/1).

References

1. Leon DA, McCambridge J (2006) Liver cirrhosis mortality rates in Britain from 1950 to 2002: an analysis of routine data. Lancet 367:52–56

2. Delco F, Tchambaz L, Schlienger R, Drewe J, Krahenbuhl S (2005) Dose adjustment in patients with liver disease. Drug Saf 28:529–545

3. Zuckerman MJ, Menzies IS, Ho H, Gregory GG, Casner NA, Crane RS et al (2004) Assessment of intestinal permeability and absorption in cirrhotic patients with ascites using combined sugar probes. Dig Dis Sci 49:621–626

4. Blaschke TF, Rubin PC (1979) Hepatic first-pass metabolism in liver disease. Clin Pharmacokinet 4:423–432

5. Vyas K, Gala B, Sawant P, Das HS, Kulhalli PM, Mahajan SS (2002) Assessment of portal hemodynamics by ultrasound color Doppler and laser Doppler velocimetry in liver cirrhosis. Indian J Gastroenterol 21:176–178

6. Adedoyin A, Arns PA, Richards WO, Wilkinson GR, Branch RA (1998) Selective effect of liver disease on the activities of specific metabolizing enzymes: investigation of cytochromes P450 2C19 and 2D6. Clin Pharmacol Ther 64:8–17

7. George J, Murray M, Byth K, Farrell GC (1995) Differential alterations of cytochrome P450 proteins in livers from patients with severe chronic liver disease. Hepatology 21:120–128 8. Tegeder I, Lotsch J, Geisslinger G (1999) Pharmacokinetics of

opioids in liver disease. Clin Pharmacokinet 37:17–40

9. MacGilchrist AJ, Birnie GG, Cook A, Scobie G, Murray T, Watkinson G et al (1986) Pharmacokinetics and pharmacodynamics of intravenous midazolam in patients with severe alcoholic cirrhosis. Gut 27:190–195

10. Gines P, Arrovo V, Rodes J (1992) Pharmacotherapy of ascites associated with cirrhosis. Drugs 43:316–332

11. Krahenbuhl-Melcher A, Schlienger R, Lampert M, Haschke M, Drewe J, Krahenbuhl S (2007) Drug-related problems in hospitals: a review of the recent literature. Drug Saf 30:379–407

12. Cockcroft DW, Gault MH (1976) Prediction of creatinine clearance from serum creatinine. Nephron 16:31–41

13. Child CG, Turcotte JG (1964) Surgery and portal hypertension. In: Child C (ed) The liver and portal hypertension. Saunders, Philadelphia, pp 50–64

14. Egger SS, Meier S, Leu C, Christen S, Gratwohl A, Krahenbuhl S et al (2010) Drug interactions and adverse events associated with antimycotic drugs used for invasive aspergillosis in hematopoietic SCT. Bone Marrow Transplant 45:1197–1203

15. Lucena MI, Andrade RJ, Tognoni G, Hidalgo R, De La Cuesta FS (2002) Multicenter hospital study on prescribing patterns for prophylaxis and treatment of complications of cirrhosis. Eur J Clin Pharmacol 58:435–440

16. Lucena MI, Andrade RJ, Tognoni G, Hidalgo R, Sanchez de la Cuesta F (2003) Drug use for non-hepatic associated conditions in patients with liver cirrhosis. Eur J Clin Pharmacol 59:71–76

17. Lazarou J, Pomeranz BH, Corey PN (1998) Incidence of adverse drug reactions in hospitalized patients: a meta-analysis of prospective studies. JAMA 279:1200–1205

18. van der Hooft CS, Dieleman JP, Siemes C, Aarnoudse AJ, Verhamme KM, Stricker BH et al (2008) Adverse drug reaction-related hospitalisations: a population-based cohort study. Pharma-coepidemiol Drug Saf 17:365–371

19. Pirmohamed M, James S, Meakin S, Green C, Scott AK, Walley TJ et al (2004) Adverse drug reactions as cause of admission to hospital: prospective analysis of 18 820 patients. BMJ 329:15–19

20. Fattinger K, Roos M, Vergeres P, Holenstein C, Kind B, Masche U et al (2000) Epidemiology of drug exposure and adverse drug reactions in two Swiss departments of internal medicine. Br J Clin Pharmacol 49:158–167

21. Classen DC, Pestotnik SL, Evans RS, Burke JP (1991) Comput-erized surveillance of adverse drug events in hospital patients. JAMA 266:2847–2851

(10)

22. Herr RD, Caravati EM, Tyler LS, Iorg E, Linscott MS (1992) Prospective evaluation of adverse drug interactions in the emergency department. Ann Emerg Med 21:1331–1336 23. Ratz Bravo AE, Tchambaz L, Krahenbuhl-Melcher A, Hess L,

Schlienger RG, Krahenbuhl S (2005) Prevalence of potentially severe drug-drug interactions in ambulatory patients with dyslipi-daemia receiving HMG-CoA reductase inhibitor therapy. Drug Saf 28:263–275

24. Corsonello A, Pedone C, Corica F, Mussi C, Carbonin P, Antonelli Incalzi R et al (2005) Concealed renal insufficiency and adverse drug reactions in elderly hospitalized patients. Arch Intern Med 165:790–795

25. Papadakis MA, Arieff AI (1987) Unpredictability of clinical evaluation of renal function in cirrhosis. Prospective study. Am J Med 82:945–952

26. Angeli P, Gatta A, Caregaro L, Menon F, Sacerdoti D, Merkel C et al (1990) Tubular site of renal sodium retention in ascitic

liver cirrhosis evaluated by lithium clearance. Eur J Clin Invest 20:111–117

27. Amir O, Hassan Y, Sarriff A, Awaisu A, Abd Aziz N, Ismail O (2009) Incidence of risk factors for developing hyperkalemia when using ACE inhibitors in cardiovascular diseases. Pharm World Sci 31:387–393

28. Sacerdoti D, Bolognesi M, Merkel C, Angeli P, Gatta A (1993) Renal vasoconstriction in cirrhosis evaluated by duplex Doppler ultrasonography. Hepatology 17:219–224

29. Wensing G, Lotterer E, Link I, Hahn EG, Fleig WE (1997) Urinary sodium balance in patients with cirrhosis: relationship to quantitative parameters of liver function. Hepatology 26:1149– 1155

30. De Ledinghen V, Heresbach D, Fourdan O, Bernard P, Liebaert-Bories MP, Nousbaum JB et al (1999) Anti-inflammatory drugs and variceal bleeding: a case-control study. Gut 44:270–273

Figure

Table 1 Drugs at hospital admission for patients with liver cirrhosis (n =400 patients)
Table 3 Prevalence of ADRs with a definite, probable, or possible causality rating in 400 patients with liver cirrhosis
Table 4 Cirrhotic patients with one or more ADRs in comparison with cirrhotic patients without an ADR

Références

Documents relatifs

Après deux heures devant la fenêtre, je voulus pour la première fois dans la journée parler à Mathias.. Mathias, c’est

[r]

RELEVANT PHARMACOKINETIC AND PHARMACODYNAMIC ELEMENTS ABOUT DRUGS CURRENTLY PROPOSED FOR COVID-19 TREATMENT The drugs currently offered for the treatment of COVID-19 are either

Seniors who take an average of five prescription medica- tions or self-care products (2.8 medications plus 1.9 self- care products) concurrently are at low risk of clinically

Health care professionals must be aware of the potential for serious adverse reac- tions, including neuropathy, when linezolid is used beyond its

If either authors or drug editorial boards share DIKB-compatible micropub- lications, other groups who need to process the same evidence would also benefit because they would not

Values of plateau storage modulus as a function of volume fraction with emulsion as square symbol and suspension as inverse triangle

SNOMED CT is the most comprehensive ontology of clinical terms for supporting health information systems [2], including different classifications of drugs and drug-